Fine-Tune it Like I'm Five: Supporting Medical Domain Experts in Training NER Models Using Cloud, LLM, and Auto Fine-Tuning

Abstract

This paper presents a system to manage, train, and optimize Named Entity Recognition models using Cloud resources, named CRM4NER. This system explores features to perform model fine-tuning using automatic parameter searches and context-aware fine-tuning recommendations using a Large Language Model. A goal is to support domain experts in the difficult task of training and fine-tuning Machine Learning models, which often requires expertise that domain experts are lacking. The system also includes functionalities to manage models and training data for the entire usage life-cycle and is integrated in a knowledge management system. Through providing Cloud-based storage and training, domain users are further supported though scalable compute and storage for advanced Machine Learning workloads such as GPU-based Transformer models for improved performance.

Ähnliche Arbeiten